Factors that impact the loan approval for an applicant

Investigation Overview:

This project is about the factors or features that could help the loan applicant to get the approval status.

Dataset Overview:

This data set contains 113,937 loans with 81 variables on each loan, including loan amount, borrower rate (or interest rate), current loan status, borrower income, and many others.

Conclusion:

In the loan_data dataset, there have 113937 rows and 24 columns.All the data types are correct.There have missing values in the EstimatedEffectiveYield,BorrowerAPR , ProsperRating (numeric),ProsperRating (Alpha) , ProsperScore , EmploymentStatus , Occupation ,EmploymentStatusDuration,DebtToIncomeRatio,BorrowerState columns.

As this project is all about visualization, for simplicity purpose null values are deleted.

Univariate Exploration

LoanStatus of all Borrowers

Conclusion:

LoanStatus of all Borrowers are with current and completed state.

EmploymentStatus of all Borrowers

Conclusion:

EmploymentStatus of all Borrowers are with Employed State and most of them are full time worker.

IncomeRange of all Borrowers

Conclusion:

People having middle middle income(50,000-74,999 USD) and low middle income (25,000-49,999 USD) tool more loans.Job less and low income people have less chance to get loans from bank.

Top 10 states of all Borrowers

Conclusion:

Top 5 states of all Borrowers are from CA,NY,TX,FL and IL

Top 10 - Occupation of all Borrowers

Conclusion:

Most of the borrowers occupation are not defined.May be self employed like property owner.But majority are with an occupation of Professional and Executive.

ProsperScore of all Borrowers

Conclusion:

Majority of the borrowers are with a rating or score from 4 to 8.They have higher chance to approve loan.

Numeric attributes analyze :

The variables that are numeric are 'term', 'estimatedeffectiveyield', 'borrowerapr', 'borrowerrate', 'prosperrating_numeric', 'prosperscore', 'listingcategory_numeric', 'employmentstatusduration', 'statedmonthlyincome', 'monthlyloanpayment', 'recommendations', 'debtToIncomeratio', 'loanoriginalamount', 'percentfunded', 'investors'

Insights with Numerical variables analysis:

BorrowerRate : The average interest rate is 0.19.The maximum interest rate is 0.36 and minimum rate is 0.04.But most of the borrowers inetrest rate is 0.16.

StatedMonthlyIncome : The average monthly income is approx. 6002 USD.The maximum income is 483333 USD and minimum rate is 0.25.But most of the borrowers monthky incomw 4500 USD.

LoanOriginalAmount: The average amount is 9294 USD.The maximum loan is 35000 USD and minimum is 1000.But most of the borrowers loan amaount is approx. 4500 USD.

Employment status duration: The average amount is 104.5 months.The maximum is 755 months and the minimum is 0.But most of the borrowers took loan whose have employement status 0-50 months.

Bivariate Exploration:

Relationship between two numeric variables

*Conclusion:

  1. Loan original amount and monthly loan payment is highly correlated.
  2. Borrower annual percentage rate and prosper score is negatively correlated.

Multivariate Exploration

Applicants rating with monthly income and Employment Status

Conclusion:

  1. For Applicants(employed and fulltime) with prosper ratings from 7 to 4 have the higher loan amount with increased salary.
  2. For Applicants(parttime employee) with prosper ratings from 7 to 4 have the lower loan amount with low level salary.

Did homeowner status has impacted the ProsperRating and Borrower Interest Rate?

Conclusion:

We observe that without homeowner tend to have a higher interest rate, and thus lower rating.However homeowner tends to have lower interest rate and higher rating. So we can safely say that homeowner is safest bet when giving a loan.

Overall findings:

To summarize this report, I believe that the loan approval status is heavily influenced by the applicant's details on income range, house owner status, and job status.